Notice and Invitation
Oral Presentation of Dissertation Proposal
Department of Bioengineering, George Mason University
Bachelor of Engineering, Electrical Engineering, University of Pune, 2014
Master of Science, Electrical and Computer Engineering, University of California Santa Barbara, 2017
An Investigation of Multisensory Feedback Integration for
Monday, September 28, 2020, 11:00am-1:00 pm
All are invited to attend.
Dr. Siddhartha Sikdar, Dissertation Director
Dr. Parag Chitnis, Committee Chair
Dr. James Thompson
Dr. Wilsaan Joiner
Current commercially available prosthetic control technologies use surface electromyography to control the opening or closing speed of the prosthetic hand. Surface electromyography has limitations with respect to its signal to noise ratio, cross-talk between adjacent sensors, etc., which necessitates the use of non-intuitive control paradigms such as velocity control. This is problematic when performing activities of daily living, resulting in a difficulty to control the prosthetic as well as a higher cognitive load. Additionally, there is a lack of multi-sensory feedback from the prosthetic hand, which forces the user to fixate their gaze on the prosthetic hand during task performance. To mitigate some of these problems, our lab has developed a novel prosthetic control technique called 'Sonomyography', that uses ultrasound cross-sections of the users' arm to deduce the intended movement.
First, this dissertation investigated if the able bodied controls as well as individuals with limb loss could reliably control an on-screen cursor using Sonomyography. Two control modalities - position control and velocity control, were used to control a virtual cursor and the users' performance of the users was quantified using their 'time to target' and 'task completion percentage'. Each control modality was found to offer unique advantages in terms of task performance. Further studies will refine the implementation of reliable proportional control using this knowledge.
Users integrate multi-sensory feedback during prosthetic control, just like able bodied subjects. Ideally, a prosthetic arm that can be controlled intuitively and dexterously should enable the user to integrate multi-sensory feedback, without increasing their cognitive load. To investigate how users do this while using Sonomyography, the effect of visual, proprioceptive, haptic feedback was evaluated by depriving users of these feedback modalities during each trial. The insights gained from these studies will inform the design of optimal feedback strategies during prosthetic design. Further studies will investigate the role of multi-sensory feedback during performance of functional tasks.
Finally, this dissertation will investigate if currently available outcome metrics and interventions can be delivered in a user-centric manner. Deprivation of each feedback modality may have a different effect on every user. This will manifest itself non-uniformly across several outcome metrics. The optimal solution may be to analyze the needs of every user individually. This user-centric needs assessment should inform the outcome metrics to be measured, and the subsequent interventions to be delivered. If future prosthetic devices are made more intuitive, multi-sensory feedback is optimally designed, and the outcome metrics and interventions are designed based on each user's needs, then it will benefit the users and increase the retention rate of prosthetic arms among individuals with limb loss.
Join Zoom Meeting
Meeting ID: 985 8819 2331